Chalice AI rebrands as custom bidding shifts toward agentic AI models

The pivot from static algorithms to generative, agentic bidding signals a new phase in programmatic maturity where brands prioritize proprietary business logic over generic DSP optimization. For the streaming ecosystem, this means high-value CTV inventory can now be priced against specific offline outcomes—like pharmacy sales—rather than just completion rates. As walled gardens face antitrust pressure, portable, containerized AI models allow advertisers to maintain consistent decisioning across fragmented supply paths. Watch for the adoption of the IAB Tech Lab’s Agentic Real-Time Framework (ARTF) as the benchmark for this interoperable bidding layer.

The move toward specialized bidding coincides with significant infrastructure updates across major demand-side platforms. Per Amazon, April 2025, the Amazon DSP launched advanced bid adjustments and a new API that allows advertisers to apply multi-dimensional bidding logic within a single line item, significantly reducing operational overhead for brands like Coty. This technical opening has enabled third-party AI firms to layer proprietary data directly onto retail signals, moving beyond the "one-size-fits-all" models that previously dominated retail media networks. Competitive pressure is also mounting from independent platforms. According to xpon.ai, May 2026, The Trade Desk’s Kokai platform has evolved to feature "Adaptive Trading Modes" that automate bidding against real-time business returns, while Google’s DV360 has integrated Gemini-powered workflows to handle complex QA and audience generation. These developments reflect a broader industry shift where AI is no longer just a feature of the bidding engine but is being integrated into the entire campaign lifecycle to manage fragmented signal loss from the phase-out of third-party cookies. Furthermore, the integration of these models into cloud environments is accelerating. Per Media-Marketing, November 2025, Bayer has already moved its programmatic optimization into Snowflake via Chalice AI, creating a composable retail media architecture. This setup allows the life sciences giant to refine its bidding models internally, ensuring that sensitive first-party data remains within its own secure perimeter while still influencing real-time auctions across the open web and Connected TV.